Randomizing blocks in a nested timeline

[j22tis]

Hi!

My goal is to build an experiment with two kinds of blocks of 15 trials with different rates of trial difficulty (e.g a hard block with 75% hard and 25% medium trials and an easy block with 75% easy and 25% medium trials, trial sequences randomized within blocks).

My approach was the following:

1. Define a general trial sequence

 experimental_trial = {
     timeline: [prosp_confidence, fixation, rdk, retro_confidence, affect_slider, experimental_feedback, iti,]
    } 

2. Define hard block

var hard_params = [{direction:"0", coherence:"0.4"}, {direction:"180",coherence:"0.4"}, {direction:"0",coherence:"0.2"}, {direction:"180",coherence:"0.2"}]

var hard_block = {
         timeline: [experimental_trial],
         timeline_variables: hard_params,
         sample: {
            type: 'with-replacement',
            size: 15, 
            weights: [0.25,0.25,0.75,0.75], 
        },
         on_start: function() {
         block_type = 'hard_block'
         phase = 'experimental'}
       } 

3. Define easy block

var easy_params = [
{direction:"0", coherence:"0.4"},
{direction:"180", coherence:"0.4"},
{direction:"0", coherence:"0.8"},
{direction:"180", coherence:"0.8"}]


// build an easy block
var easy_block = {
     timeline: [experimental_trial],
     timeline_variables: easy_params,
     sample: {
        type: 'with-replacement',
        size: 3, // 10 trials, with replacement
        weights: [0.25,0.25,0.75,0.75], // The Alex trial is three times as likely to be sampled as the others.
    },
     on_start: function() {
     block_type = 'easy_block'
     phase = 'experimental'}
    }

4. Try to combine the hard and easy blocks into a new timeline with something like this:

var blocks = ["hard_block", "easy_block"];

var experimental_timeline = jsPsych.randomization.sampleWithReplacement(blocks, 6);

experimental_procedure = {
     timeline: experimental_timeline} 

However, the last part (4) does not seem to work and I have not found a way to randomize my hard and easy blocks in the experimental procedure level. So I'm wondering is my approach even possible or should I try to tackle it some other way?

[becky-gilbert]

Hi @j22tis, I like this approach and I think it should work. I think the reason your last step isn't working is because you're sampling from an array of strings, rather than sampling the actual objects that are named with those strings. So try doing it without the quotes around hard_block and easy_block:

var blocks = [hard_block, easy_block];

var experimental_timeline = jsPsych.randomization.sampleWithReplacement(blocks, 6);

experimental_procedure = {
     timeline: experimental_timeline
} 

Does that work?

[becky-gilbert]

Also, maybe add a console.log just to check that the output of the randomization function is what you're expecting, i.e. an array of 6 easy_block and hard_block objects:

var experimental_timeline = jsPsych.randomization.sampleWithReplacement(blocks, 6);
console.log(experimental_timeline);

[j22tis]

Yes, this works! I actually managed to figure it out, but didn't get to updating here :( Still glad that you are suggesting the same thing :)

Many thanks,
Alan

[becky-gilbert]

Sorry that my response wasn't quick enough to be helpful, and glad to hear that you figured it out!